A New Class of Upper Bounds on the Log Partition Function
Martin J. WainwrightTommi S. JaakkolaAlan S. WillskyPublished in: CoRR (2013)
Keyphrases
- partition function
- upper bound
- upper and lower bounds
- lower bound
- closed form
- belief propagation
- markov random field
- structured prediction
- importance sampling
- approximate inference
- parameter estimation
- random fields
- probability density function
- feature selection
- maximum likelihood estimation
- message passing
- bayesian framework
- learning algorithm
- sample size
- class labels
- bayesian networks
- graph cuts
- maximum likelihood
- support vector machine